Back to the blog
Reports September 19, 2026 | 5 min read

By Kareem Farid, co-founder of Kunkafa

Gold price forecast results: how far did price go? 14 to 18 September

Five days of Kunkafa’s AI market analysis and forecasts on Gold at 80% confidence and above, every forecast measured by how far the gold price actually went toward the level it named.

Every Gold forecast from Kunkafa’s AI market analysis names a direction, a duration and a level. This weekly report takes every gold price forecast Kunkafa published from Monday 14 to Friday 18 September at 80% confidence and above, and asks one question of each: inside its own duration, how far did price actually go toward that level?

Gold bars on a dark aubergine background
16.8%
went at least twice as far as the level
66.0%
reached the level, up to twice its distance
17.2%
stopped short of the level
       
3× and morelevelhalfwayshort
7.5%3×+9.3%2×–3×66.0%reached4.1%75–99%3.9%50–75%4.1%25–50%5.1%<25%

60,858 forecasts at Kunkafa’s confidence of 80% or higher, both directions, every duration up to a day.

Each forecast names a level. This measures the furthest price travelled toward it, or past it, inside the forecast’s own duration, divided by the distance to the level. 100% is the level itself; 300% means price travelled three times as far.

Confidence, and what it was worth

Every forecast carries Kunkafa’s confidence. The higher the confidence, the further price tends to go. That is the whole point of publishing it. The opening picture pools everything from 80% up; here it is split into its two bands.

Confidence 90% and above22,156 forecasts
       
10.1%3×+11.0%2×–3×65.3%reached3.2%75–99%3.1%50–75%3.1%25–50%4.2%<25%
Confidence 80 to 90%38,702 forecasts
       
6.0%3×+8.4%2×–3×66.4%reached4.6%75–99%4.3%50–75%4.7%25–50%5.6%<25%

Both directions

A rise is never shown without the fall beside it. Here is each side of the five days.

▲ Could rise30,536 forecasts
       
7.4%3×+9.5%2×–3×67.7%reached3.8%75–99%3.8%50–75%3.6%25–50%4.2%<25%
▼ Could fall30,322 forecasts
       
7.6%3×+9.2%2×–3×64.3%reached4.3%75–99%4.0%50–75%4.6%25–50%6.0%<25%

By duration

Kunkafa answers the same question at many durations, from minutes to a day. Each forecast is judged only within its own duration.

Up to 5 minutes11,046 forecasts
       
25.6%3×+20.0%2×–3×36.7%reached4.1%75–99%3.6%50–75%4.1%25–50%5.9%<25%
Up to 25 minutes12,864 forecasts
       
8.1%3×+14.9%2×–3×60.4%reached4.0%75–99%4.0%50–75%3.8%25–50%4.8%<25%
Up to 1 hour11,319 forecasts
       
2.7%3×+7.4%2×–3×73.0%reached4.0%75–99%3.9%50–75%4.1%25–50%4.9%<25%
Up to 2 hours2,076 forecasts
       
2.8%3×+10.7%2×–3×66.9%reached5.3%75–99%4.2%50–75%4.6%25–50%5.5%<25%
Up to 6 hours12,729 forecasts
       
0.7%3×+2.4%2×–3×80.3%reached4.1%75–99%4.0%50–75%3.8%25–50%4.7%<25%
Up to 1 day10,824 forecasts
       
2.4%3×+1.8%2×–3×78.1%reached3.9%75–99%3.8%50–75%4.8%25–50%5.2%<25%

Only settled forecasts are counted: the full duration had elapsed, and for durations beyond two days it had elapsed at least a day before 18 September. Nothing longer than a day has settled yet under that rule.

Day by day

The same breakdown for each day’s forecasts, Monday to Friday.

Mon 14
3,237
Tue 15
15,469
Wed 16
7,756
Thu 17
18,489
Fri 18
15,907
3× the level or more2× to 3× the levelReached it, up to 2×75 to 99% of the way50 to 75% of the way25 to 50% of the wayUnder 25% of the way

What the numbers look like, one forecast at a time

Real forecasts from the week, confidence 80% and above, durations of an hour or more: in each band, the one with the widest distance to its level. Prices in USD per ounce.

3× the level or more · 4,581 forecasts

▲ Rise · Up to 1 day · 2026-09-15 20:40 UTC
Entry
4,293.14
Level
4,305.85
Distance
0.30%
Furthest move
1.12%
Confidence
85%
378% of the way · went past the level

2× to 3× the level · 5,685 forecasts

▼ Fall · Up to 1 day · 2026-09-16 13:45 UTC
Entry
4,343.00
Level
4,328.53
Distance
0.33%
Furthest move
0.67%
Confidence
81%
201% of the way · went past the level

Reached it, up to 2× · 40,155 forecasts

▲ Rise · Up to 1 day · 2026-09-17 09:59 UTC
Entry
4,306.39
Level
4,327.73
Distance
0.50%
Furthest move
0.52%
Confidence
82%
106% of the way · went past the level

75 to 99% of the way · 2,474 forecasts

▼ Fall · Up to 1 day · 2026-09-17 01:13 UTC
Entry
4,298.01
Level
4,271.12
Distance
0.63%
Furthest move
0.56%
Confidence
81%
88% of the way · stopped short

50 to 75% of the way · 2,362 forecasts

▲ Rise · Up to 1 day · 2026-09-16 13:36 UTC
Entry
4,347.89
Level
4,362.76
Distance
0.34%
Furthest move
0.18%
Confidence
80%
53% of the way · stopped short

25 to 50% of the way · 2,510 forecasts

▼ Fall · Up to 1 day · 2026-09-17 02:38 UTC
Entry
4,285.60
Level
4,247.32
Distance
0.89%
Furthest move
0.27%
Confidence
81%
30% of the way · stopped short

Under 25% of the way · 3,091 forecasts

▲ Rise · Up to 1 day · 2026-09-16 13:15 UTC
Entry
4,359.78
Level
4,374.32
Distance
0.33%
Furthest move
0.01%
Confidence
81%
4% of the way · stopped short

How this was counted

  • Source: Kunkafa’s outcome archive of live Gold forecasts, opened between 14 and 18 September 2026, at Kunkafa’s confidence of 80% or higher, both directions, every duration. Every section counts the same population.
  • Only settled forecasts are counted: the full duration had elapsed by the time of this report, and for durations beyond two days it had elapsed at least a day before 18 September.
  • “Reached” means the price touched the forecast level at any point within the forecast’s own duration. A forecast that expired without touching it counts as not reached, whatever happened afterwards.
  • No forecast was removed, re-weighted or selected by outcome.

Kunkafa’s forecasts are probabilities, not prophecy. Past results describe what happened; they do not promise what could happen next. Forecast, not advice.

Keep reading

Twelve guides that go with this one, by the question you came with.

All articles